(1 to 100 of 157 replies)
thorsten198 10:32pm, 12 December 2012
I just got aware, that the Nikon SB900 and the SB910 are on full power slower than the sync-speed, will mean you don’t get full power above 1/125 sec and this is also officially confirmed by NIKON!

The flash is also showing this reduce of power in the display, the distance range is decreasing above 1/125 sec!
The SB 800 doesn’t have this problem.

I’m pretty p… now!
Thorsten
1
(1 to 100 of 157 replies)
The SB800 has this problem too. The reduction starts at 1/60s, although it might not be reflected in the LCD.
Rangefindergeneral 5 years ago
would you care to site your sources..
admin
strobist PRO 5 years ago
It's true - the full-power t.1 times of many speedlights exceeds a 250th of a sec. At a 250th on an SB-800, you get about 1/2 a stop more out of it going from 1/2 to 1/1 due to the lopped off pulse.
Rangefindergeneral 5 years ago
strobist:
I've not noticed this in my images, I'll need to look in to this.. I shoot a lot at 1/1 so does this mean I might be better served at 1/2 power..??
thorsten198 5 years ago
I’ve not noticed this either, but think about the consequences – if the Nikon flashes are too slow for 1/1 power, than all those Yongnuos, Lumopros, whatever are also too slow for sure!
Shooting at 1/250 or 1/200 will already limit your flash power!

Which power a flash will perform is not defined by guide number, the flash duration is the main factor and nobody tells you the truth, the manufactures don’t provide the t=0.1 time, so how do we know what we buy?
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
So, let me ask this....How come that the trigger people claim full power up to 1/8000 :))

God forbid if we are limited to 1/60, how in the world are we going to overpower the sun now?
LOL

I am glad to be old enough to have grown up with a sync speed of 1/60 max. and I am pretty happy with it.

There is something peculiar about this power loss, does that mean that if I meter the flash in manual mode, the reading will be wrong?
I have not noticed that my readings produce wrong results with my Nikon flashes at 1/200.
I use my AB Einstein and AB B1600's at 1/160 and have never noticed a apparent drop in exposure when compared to the metered reading.
But I do have to admit that a half stop is barely visible in camera.

So what gives?.
ChrisVPhoto 5 years ago
Alfredk: The longer the flash duration for stuff like HyperSync, the better!
103antares 5 years ago
Not sure what you are "P..." about. It is a matter of physics and compromise. A faster flash duration would yield either lower power (which you can achieve anyway), or a color temperature that is way out of whack, or a second flash tube that adds to the fragility complexity, size and cost of the flash.

Besides, your images will not suddenly suffer due to this knowledge.
Alfredk PRO 5 years ago
I do understand that!

What I am talking about is....
Flash in manual full power, the power of the flash will not change!
Meter reads f8

Now my exposure will be OK at 1/60 but will change at 1/200 by 1/2 stop?

I am not saying that I don't believe it, I just never noticed it!
I will have to run some tests.
And there again, does 1/2 stop make really any difference?
Mr. Speedlight Posted 5 years ago. Edited by Mr. Speedlight (member) 5 years ago
...the flash duration is the main factor and nobody tells you the truth, the manufactures don’t provide the t=0.1 time, so how do we know what we buy?

But do T0.1 and T0.5 mean very much with a truncated or quenched flash from a speedlight? The tail is cut off so I'd think from 1/2 and down the T0.1 and T0.5 would not be much different. I'd expect a transition in between 1/1 and 1/2.

SB-800 flash duration per Nikon...

1/1050 sec. at M1/1 (full) output
1/1100 sec. at M1/2 output
1/2700 sec. at M1/4 output
1/5900 sec. at M1/8 output
1/10900 sec. at M1/16 output
1/17800 sec. at M1/32 output
1/32300 sec. at M1/64 output
1/41600 sec. at M1/128 output

Note the 1/1050 at 1/1 and the 1/110 at 1/2. The 1/2 is truncated so I'd link the T0.5 and T0.1 would be almost the same while the T0.5 of the 1/1 would be notably shorter than the T0.1. These days they should give T0.5 and T0.1. Back in 1979 ASCO gave T0.5 and T0.05 in the instruction manual for my ASCO QU-4.

I'm sort of doubting the 1/2 stop loss at full power and 1/250 v. full power at1/125. I'd expect less of a loss, perhaps 1/4 stop. Does anyone have figures. Does anyone have hard proof or is 1/2 stop just a guess?

Dave Hartman

---

The SB 800 doesn’t have this problem.

From the list for the SB-800 (above) I think the SB-800 must perform in a similar manner to the SB-900/910.

Dave
Mr. Speedlight 5 years ago
Oh my God it's true! I just did a test with my D300s and SB-800. I started at 1/320th, f/16 at full power on a white wall and watched the spike march to the right as I dropped the shutter speed by 1/3rds down to 1/60th. Isn't physics a mother raper!

Most of the action was between 1/320th and 1/200th with the spike sort of marching in place after 1/160.

Now I think we Nikon boys should observe a moment of silence. If we have to drop to a 1/125 to get full power the Canon boys are going to have to drop to 1/60 or even lower. Oh the humanity!

Dave Hartman
Mr. Speedlight Posted 5 years ago. Edited by Mr. Speedlight (member) 5 years ago
At half power, again with the SB-800 and D300s there is a marked loss at 1/320th to a 1/250th but from a 1/250th down to 1/60th the spike on the histogram just marches in place.

Dave
Does this mean that the D600 with its apparent handicap of a 1/200 flash sync speed will not give you the quite the drop in effective flash power as discussed?

1/250 to 1/200 is what, 1/3 stop? And how much do we lose (for a SB800/900) at full power at this speed, at least 1/4?

Is it possible that the D600 isn't actually handicapped at all.....?

;)

would you care to site your sources..

Its a well known fact. (Perhaps not?)

If your shutter speed is shorter than your t.01 times you cut output. The reduction in output compared to the increase in shutter speed up to your maximum x-sync doesn't have a negative impact as usually the increase in shutter speed (and reduction in ambient) is not as great as the loss in flash output, so the optimal setting for using flash mixed with ambient is usually at the maximum x-sync. Digital shutters can push this benefit beyond x-sync but only to a point.

You do get an increase in output at slower shutter speeds, so in the studio without ambient to contend with, lower shutter speeds provide a greater flash benefit.

Heres an interesting comparison from 2006 illustrating the loss in output from an SB800 in relation to shutter speed increase and the comparative supremacy of a digital shutter with the same SB800:
SB800 Output comparison

I won't mention the NEED for multiple Speedlights in this respect - but I will point out the LOSS :-)
aperture-priority 5 years ago
Because of the way speedlights truncate the flash output below full power, this is only really an issue when using full power. Others have explained the detail above.

I would add that if you are working with a speedlight at full power regularly, then the biggest issue will probably be the longer recycle times and heat build up problems. Using a second light bungeed to the first and running both at 1/2 or less power gives the same output without suffering the light loss and the slow recycle times.

YMMV
Bill.
Bobsden 5 years ago
Does a tree falling in the woods make a noise if there is no-one to hear it?
Rangefindergeneral 5 years ago
Commercial Photographer:
Useful but these tests are for FP shutter.
I will run my own tests when I have some time.
Thanks
LayerMask Posted 5 years ago. Edited by LayerMask (member) 5 years ago
This 1/1 M tail clipping seems pretty universal when manufacturers are coaxing the most out of AA batteries.

Canon's 580ExII is the same. On a body with a max sync of 1/250th, speedlite at 1/1 M, there's very roughly 1/4 stop gain going from 1/250th to 1/200th. Edit: That's with the speedlite in the cam's hot-shoe - with its light bouncing off a low ceiling. Possibly a configuration that emphasizes the power difference.

As expected, just going to 1/3rd stop below 1/1, then this issue disappears.

On 1/160 bodies, this is a non-problem - although they have others, of course, if trying to balance bright ambient etcetc.
mjkzz (a.k.a zwdeal) 5 years ago
aperture-priority:

+1

Despite what manufacturers say in the manual, at full power, the tail is long... for my SB600 (bought around 2005) at full power, T.0 (not T.1 nor T.5) is ~8.2ms . . . sure it is an old unit, caps might not be as good as before, but still.

But my SB600 at half power is much faster, T.0 is around 1ms, generates less heat.
Alfredk PRO 5 years ago
So the question remains, does a 1/4 to 1/2 stop actually make a big difference in a practical application, I would say, not to me, otherwise I would have noticed it before.
Rangefinder general:

The tests were from 1/60s and clearly illustrate the reduction of output from there. No FP mode involved.

FP mode doesn't start until 1/250s on the D200 and the Digital shutter on the D70 doesn't use FP mode at all. The digital shutter only commences from 1/500s. The digital shutter results will illustrate the amount of output capable of being captured during the duration of the shutter speed illustrated as the sensor simply stops exposing and cuts off collecting light at anything longer. This illustrates the comparative reduction you were looking for.
Jerry P. H. Posted 5 years ago. Edited by Jerry P. H. (member) 5 years ago
Interesting. Stuff like this I like to test, maybe I will this weekend.

Would a Sekonic light meter also show these drop-offs in power?

Edit: even then, a 1/2 stop is no biggie and never stopped me from getting the shot in the real world.
LayerMask Posted 5 years ago. Edited by LayerMask (member) 5 years ago
" .. Would a Sekonic light meter also show these drop-offs in power? .. "

Maybe not - it's not that the Speedlight isn't delivering - it's just that it's not delivering quickly enough. Quickly enough for a shutter at its Max sync. that is - your meter will be more patient. Please let us know.

My guesstimate above was just from oggling the histo.

Edit: Maybe the M 1/1 clipped tail, shutter at Max sync, would be pretty orange anyway??
Rangefindergeneral Posted 5 years ago. Edited by Rangefindergeneral (member) 5 years ago
As I shoot at 1/1 a lot I shall have to do some testing because it would be fine with me to drop the power to 1/2 with less loss. But I still seem to find I need the extra ummph from the 1/1 when outside, although I will now think more about it and when to use it.. I'll be dusting off the D70...
Thanks for the info CP and others..
tundracamper 5 years ago
Then folks need to quit recommending Manual Settings of the sync speed (1/250) and then adjusting the aperture for the flash or DOF you want. If 1/60 and 1/250 will both accomplish the same thing in terms of ambient, then obviously it's better to use 1/60 with the gain in flash power due solely to the physics. Good to know!!

Uh, including Mr. Hobby!
"Since the light from your flash is pretty much instantaneous, it really does not care about the shutter—as long as you are at or below your camera's top "sync" speed. Which for most cameras is either 1/250th or 1/200th of a second. "
strobist.blogspot.com/2006/03/lighting-101-balancing-flas...

Actually, it does care about the shutter, then, doesn't it? (This is not directed solely at him, as I have seen this generic recommendation many, many, places. It just needs a footnote about decreased flash output at full power.)
Alfredk PRO 5 years ago
Well I run some tests with my Einstein at full power and I do not see the spike in the histogram to move on any setting between 1/60 and 1/200, I will try the SB's next as soon as I have a chance.
Mr. Speedlight Posted 5 years ago. Edited by Mr. Speedlight (member) 5 years ago
Useful but these tests are for FP shutter.
I will run my own tests when I have some time.
Thanks


I wish there was an easy way to show the histograms for my quick test above. Even though I knew this was happening there is a more pronounced loss of power than I thought there would be. It's quite pronounced at 1/320 even though my D300s doesn't switch to Auto FP until 1/400.

Its a well known fact. (Perhaps not?)

I think many know this but then Blithely over look it. That was the case with me. If I could see the tail on an oscilloscope of a full power flash from my SB-800 it must be longer than I thought it would be. I thought the loss at 1/1 would be something like 1/4 stop. My D300s will give me clean 1/320 x-sync without Auto FP (HSS) but even without Auto FP the loss at 1/320 is pretty steep. My D300s doesn't switch to Auto FP until 1/400 second with an SB-800 or 700.

...Would a Sekonic light meter also show these drop-offs in power?...

I'm wondering this also. I'm still using an old Minolta Flash Meter III and I don't think it can. I would hope newer flash meters will.

Using a second light bungeed to the first and running both at 1/2 or less power gives the same output without suffering the light loss and the slow recycle times.

I seldom use a speedlight at full power for just these reasons. I want faster recycling so I can take a second or third shot in quick succession. If in manual I'd prefer to use two speedlights at 1/2 power than one at 1/1. If in TTL I'd prefer to bump the ISO up a stop or so.

Again I thought the loss would be less. Now that the issue is brought to focus I know I've seen it in tests and wondered briefly about the visual loss on the LCD then blithely ignored it. As mentioned above if the ambient light is low and one is bouncing or reach out to a distant subject then dropping the shutter speed to 1/160 might be useful.

Dave Hartman
Mr. Speedlight 5 years ago
I'll be dusting off the D70...

But will the D70 help? Won't it's electronic shutter truncate the flash tail just like a conventional shutter. If using a single speedlight with a DSLR I think the best bet is bumping the ISO a stop and or dropping the shutter speed one or two thirds stops.

Dave
John Adkins II PRO 5 years ago
Ok, not quite sure I'm following the discussion correctly, but it seems that you guys are saying a Nikon SB-910 / 900 / 800 at full power, will lose an actual stop/ 1/2 stop of power going from say 1/125 to 1/250 sec?

I've been using these flashes a long time and can honestly say I've never noticed this before, so I decided to test it myself if this is indeed what's being discussed.

First off I completely understand the loss of light when you exceed the camera's sync speed, such as going from 1/250 to 1/500.

So for my test (not scientific at all really), I parked an SB-800 (I don't have the other two) in a Lastolite Ezybox softbox camera left, cranked it up to full power and took 4 shots, walking the shutter speed up from 1/60 to 1/125 to 1/250 to 1/500 and the only noticeable difference I could see is from 1/250 to 1/500.

Light Loss Test

Maybe this isn't what's being discussed, or maybe there is a loss that my eye just doesn't see, but in any case, if this is in fact the loss you're talking about, I see absolutely no real reason to be concerned about it.

Again, I just did this for my own knowledge, but if anyone has anything to add, or correct me on, please do! (like I had to ask) ;)
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
I am coming up with the exact same results, so I am not sure where this comes from?

I see no visual or measurable difference between 1/60 to 1/200 with my Einstein, SB700 and SB800 on full power.
NADA

I do know from earlier tests that I start losing exposure when using my AB B1600 on low power and a speed of 1/180 and above.

I did all tests with my D7000
thorsten198 5 years ago
Thank's John
John Adkins II PRO 5 years ago
Alfredk:

Maybe it's just such a slight difference that most people just don't see it... or it's only explainable with 'pseudo-math' ;)

thorsten198:

You're welcome! Although I'm not even sure I'm following this discussion 100%. :)
mjkzz (a.k.a zwdeal) 5 years ago
John Adkins II:

Thanks for the tests.

I think your tests illustrate the issue well -- if full power of SB-800 is what Nikon claims to be, at 1/500 shutter speed, it should not have much difference because Nikno claim 1/1 is 1/1050 per Mr. Speedlight. Of course, it could be sync issue.

For 1/60 one, the shutter stays open for 16.67ms, long enough to cover all flash duration. For 1/125, that is 8ms, again should be long enough to cover flash duration, even at T.0 probably.

For 1/250, that is 4ms, I am not sure about this, but if you pay attention to model's right hand, I think it is a bit darker and skin tone is more saturated than the previous two shots.

Flash tail has much less intensity than the peak, so 4ms might not cut a lot of power, so maybe a meter can tell.
mjkzz (a.k.a zwdeal) 5 years ago
LayerMask:

I think you can set shutter speed on Sekonic light meter.
thorsten198 Posted 5 years ago. Edited by thorsten198 (member) 5 years ago
So I guess this will prove that the SB800 doesn’t have the problem, the SB700 won’t have it then either since it’s faster – this would confirm what I’ve heard and what I wrote in my opening post.

Yesterday night, when I got aware of this due to a discussion Metz 50 vs Nikon SB 700 (Metz is listing 1/125 sec flash duration at 1/1) I took the D300 with 50 mm lens attached, clicked the SB 800 in the hot shoe and dialled up shutter speed to 1/250 (ISO 200 f/2.8) the distance range on the LCD was consistent at 20m, then I repeated the procedure with the SB 900 and the distance range dropped at 1/200 to 19 metres, at 1/250 to 18 metres.

I guess John’n test confirmed, that the SB 800 doesn’t have this problem (that’s what I’ve heard and wrote in my opening), so the SB 700 won’t have this problem either since it is faster.
I haven’t done a test with the SB 900 yet, but I try to repeat John’s test tomorrow with the same softbox and the SB 900.

EDIT:
Doing the test from hotshoe of the camera is different!
Both flashlights SB 800 and SB 900 are are "losing" usable power with increasing shutter speed!
thorsten198 Posted 5 years ago. Edited by thorsten198 (member) 5 years ago
I cannot see a difference in light here, that the hand is slightly brighter on one frame doesn’t mean anything, he is not a puppet, so there is also some body movement involved, look at the background. I don’t see any notable difference there.

I don't know how to meter this, the flash for the meter will be on full power at all three images for sure, but the question is, will the shutter also get full power or not?
mjkzz (a.k.a zwdeal) 5 years ago
thorsten198:

It is day time where I live so I could not test it.

I have Sekonic L-308S and it allows me to set shutter speeed even for flash measurement.

Long time ago, I did a test measuring YN560 II. At 1/125 and 1/250 speed, I do not notice much difference in reading, this is fine because YN560 has T.0 duration of 4ms. But when shutter speed is set at 1/500 on the meter, I noticed a difference of about 1/3 stop. It was beginning of the year so I do not have all data.

But since I do not have SB-800, I could not perform this test. The best thing about using meter is that we can rule out sync issue., John's tests showed more than 1/3 stop between 1/250 and 1/500, so it could be sync related.
ChrisVPhoto Posted 5 years ago. Edited by ChrisVPhoto (member) 5 years ago
1/250 isn't using FP mode, but 1/500 is. The pulsed light and shutter slit opening drops actual light-on-sensor down by comparison
mjkzz (a.k.a zwdeal) 5 years ago
ChrisVPhoto: oh, I thought the test was done with D70, the kind with electronic shutter. Sure, if SB-800 gets into FP mode, it will be different, making comparison irrelevant.
LayerMask 5 years ago
@zwdeal

" .. I think you can set shutter speed on Sekonic light meter .. "

Point taken, thanks.

But does that adjustment actually shorten the sampling time or is it just a math prediction of a shorter time effect? If Jerry PH tests, I hope he uses both his meter and his histo. Tx.
John Adkins II:

So for my test (not scientific at all really), I parked an SB-800 (I don't have the other two) in a Lastolite Ezybox softbox camera left, cranked it up to full power and took 4 shots, walking the shutter speed up from 1/60 to 1/125 to 1/250 to 1/500 and the only noticeable difference I could see is from 1/250 to 1/500.

Light Loss Test by John Adkins II


Maybe this isn't what's being discussed, or maybe there is a loss that my eye just doesn't see, but in any case, if this is in fact the loss you're talking about, I see absolutely no real reason to be concerned about it.


Your samples do show a 5% reduction - and thats not even allowing for some wireless or radio link of sorts or any cable delay.

Worse case scenario is that you're missing the peak output entirely.

The previous link I posted was the evaluation of 5 tests taking the average from each test to provide results from a hotshoe (as in the OP's case) mounted flash. Any introduction of radio, cable or CLS only deters from what the optimum result might be, but will provide a 'your case' scenario.

In your case there doesn't seem to be any worthwhile loss to talk about.
Rangefindergeneral Posted 5 years ago. Edited by Rangefindergeneral (member) 5 years ago
Argh, I'm not sure I really care....
To find out I'll have to test for myself, and I hate shooting brickwalls..
I'm in the studio tomorrow so I'll throw a few flashes on a few cameras..
If I have time...
so what do you think shoot a white brick wall take a reading with a meter, get the exposure to f16 at full power on flash then roll up the shutter speed..??? I'll stick a grey card on the wall too..
D70
D300
D700
sb800
sb900
Bowens
thorsten198 5 years ago
I just made a quick test from the hotshoe.
D300 on tripod (50mm/1.4G at f16 ISO 100)
This is only describing the histogramm:

I only started at 1/100sec wiith both flashes, but with both you will get an significant loss of power!
The SB 800 is consistant to 1/160, then there is a jump to the right at 1/200 and no visable (histogramm) difference to 1/250

The SB 900 is consitsant to 1/160sec, then a smaller loss to 1/200 and another smaller loss to 1/250

The SB 800 seems to lose far more light, than the SB900

I haven't measured anything, this is just my impression from the histogramm, but I see the reason for this thread opening confirmed.
thorsten198 5 years ago

That would be very nice, if you could do a solid test, I'm really interested in the results.
John Adkins II PRO Posted 5 years ago. Edited by John Adkins II (member) 5 years ago
thorsten198:

Couple of quick questions, are you setting the flash power manually? even while on the hotshoe of the camera?

Also, are you looking at the results of the actual test pictures you are shooting or are you just relying on data, i.e. histograms and such?

I shot my series with a D300 and an 18-200 VR lens. The pop-up flash was the triggering method using CLS, with the flash power dialed in manually at the camera.

I also didn't realize you were concerned with this issue from the flash being mounted *on* the camera and not off.

Again (and this is just my personal opinion) if there is a light loss, I just don't see it as being enough to be worried about. I do plan on testing this again with the flash on camera (yuck) just to see what happens.
I've lost my faith in lighting.

My life has been a total lie.

How did I ever manage?
Jerry P. H. Posted 5 years ago. Edited by Jerry P. H. (member) 5 years ago
John, thanks for doing the tests!

In thinking on how the test could be improved, first thing that has to be removed is most if not all ambient, that way it has no effects on the tests.

Second, like it or not, to best see any differences, a white backdrop would show things best. It is easy to measure any potential differences, even in apps like Lightroom.

Third, taking it all with a grain of salt. Flashes are not going to be all 100% the same so no matter what, the results of anyone's tests, another's may or may not be a little different.

In thinking a little more about the meters, they likely will not show this difference, but the test is worth trying as it is a fast and easy test to do. I may try this out tonight, but with further thinking on this... I use PW units, they are pretty much one of the fastest units out there and the results they give me may be quite different than if someone did the same test using say Cactus V2 units, just another factor to take into consideration.

Edit... I had a few seconds before leaving for work and did the test with the SB-900 on a stand. With the ControlTL units and the AC3set to +3 (full power), the Sekonic meter can measure a 1/3rd stop difference but under specific conditions. You have to be close enough so that the meter registers F/20 (about 4-5 feet), and then changing the shutter speed setting in the Sekonic, the reading changes from F/20 at 1/60th to F/18 at 1/250th, there is your 1/3rd of a stop drop.

If I moved the meter a little further back to where the meter read F/16, the shutter speeds made no difference, it stayed the same F/16 at either 1/60th or 1/250th.

Based on this limited and really fast and unscientific test, I would not lose any sleep over this. Later today I will do the test and see how the histogram of a D4 interprets similar results.
John Adkins II PRO Posted 5 years ago. Edited by John Adkins II (member) 5 years ago
Jerry P. H.:

Welcome,

and FWIW, there was no ambient at all contributing to those photos. f/13 and ISO 200 even at my lowest shutter speed (1/60) rendered a completely black frame.

I was also using a white backdrop, white seamless paper to be exact.

Also, the diffusion screen on my Lastolite softbox has yellowed a bit, but honestly, since I use it to light portraits (mostly) I sorta like it, instant 1/4 CTO'd.
thorsten198:
Doing the test from hotshoe of the camera is different!
Both flashlights SB 800 and SB 900 are are "losing" usable power with increasing shutter speed!


As already pointed out by Thorsten, I can't see the need to shoot any brick walls or anything.

The LCD on the SB900 tracks the reduction.. the SB800 doesn't. I take no notice of the LCD on the flash anyway and my flash meter reads output at the shutter speed I set it for if needed - and it changes appropriately.

This output loss appears to be a revelation, and must surely be another nail in the coffin of the SB800 with what must now be described as possessing a 'totally inadequate' LCD. :)

There would be some merit in testing a digital shuttered camera with various flash units to establish the optimum operating settings, but in all other cases the regular maximum x-sync will always provide the most optimal result in any high ambient situation - with whatever loss may be present.
elv0000 5 years ago
I just tried a few flashes on the camera hotshoe, and there is a very slight gradient at the bottom of the frame as you move up towards 1/250th.

You would need to shoot a while wall to see it though (which I did).

I don't understand how a light meter would give any meaningful result here, this is going into the same territory as hypersync, which bit of the massive gradient across the sensor was the meter supposed to have metered correctly? (if you see what I mean).
Jerry P. H. 5 years ago
Read my edited post, it does show a 1/3rd drop, no biggie. :-)

I will do the histogram test later on today when I get in from work.
mjkzz (a.k.a zwdeal) 5 years ago
The issue here is whether Nikon SB-900 has long tail, long enough to be cut off by a relative fast shutter so that exposure is less than what is expected.

Adding trigger in, you are going to worsen it.

Using camera to do test is not accurate enough or we human might not be able to tell the difference. Also camera has sync factor,it is going to distort result.

Use a meter and set the shutter at different speed on the meter, it will tell you if Nikon SB-900 has long tail or not and by how much.

Stick to the core, forget gradient, forget FP, shoot at the meter, not a white wall (you can't tell if exposure diff is not that much)
tundracamper 5 years ago
This seems related "SB800 Flash Duration":
www.flickr.com/photos/zelig2/3363442862/in/photostream/

SB800 Flash Duration by zelig2


Shame we don't have such curves for the SB-900.
elv0000:
I don't understand how a light meter would give any meaningful result here, this is going into the same territory as hypersync, which bit of the massive gradient across the sensor was the meter supposed to have metered correctly? (if you see what I mean).


It would be neat to think that it would be the same part (or duration) that an electronic shutter would see.. that being an absolute.. or just THAT part of a flash which would contribute in a 4s exposure in a dark room.. (ambient and flash exposures are analysed separately). BUT, all flash meters are different and where some might be set up to read a shutter speed of 1/8000s as 1/8000s, others will be set to read 1/8000s as 1/250s - nearer the truer duration of FP shutters. So it depends on the meter.
thorsten198 5 years ago


Of course the Flash was on M 1/1
No ambient in my living room due to f/16 ISO 100
I want to test the flash performance and not the trigger, so the statement of commercial photographer made sense to me, so flash on camera.

I did not made a very serious test, it won't be science anyway, so I simply wanted to know for my personal estimation, the histogramm was obvious.



It's not that I was crying all night long and have to double the anti-depressiva now, but I like to know what I have bought and I want to know what these items can do or not.
mjkzz (a.k.a zwdeal) 5 years ago
tundracamper:

this is what I expect. ... Nikon manual is probably quoting some other measurement, to get 90% power, flash duration is close to 4ms.
Jerry P. H.:
I use PW units, they are pretty much one of the fastest units out there and the results they give me may be quite different than if someone did the same test using say Cactus V2 units, just another factor to take into consideration.


What EXACTLY is the latency of your PW units?
thorsten198 5 years ago
PWs are certainly not faster than the hotshoe ?!
Bobsden 5 years ago
Don Giannatti has, after this discussion, released his lifes work is now over. He now lives in a state home for the bewildered.
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago

You said......
I only started at 1/100sec wiith both flashes, but with both you will get an significant loss of power!
The SB 800 is consistant to 1/160, then there is a jump to the right at 1/200 and no visable (histogramm) difference to 1/250


Now, having a jump to the right would mean that you gained exposure and not that you have lost it, unless your histogram is reversed!
Something is not right here!

I have a issue using the histogram for this test because a histogram is not a indication of actual exposure, it nearly shows the amount of pixels in a particular level.

John's test is the only valid test as far as I can see!

I am getting the feeling that a lot of this hype is caused by bad test procedures and trigger latency.
Alfredk:
John's test is the only valid test as far as I can see!

I am getting the feeling that a lot of this hype is caused by bad test procedures and trigger latency.


You've just stated a contradiction in terms..

.
John Adkins II PRO 5 years ago
It would be nice to see more REAL test results, i.e. images or something as opposed to opinions and ideas.
Jerry P. H. 5 years ago
I don't see it getting any more "real" than the tests you did John... LOL
Jerry P. H. 5 years ago
thorsten198 says:
PWs are certainly not faster than the hotshoe ?!
================

I doubt that they are, I did not mean to insinuate otherwise, but since this forum is about off camera flash, suddenly trigger speed is a factor that should be considered and is a factor too.

One would see greater differences with say that Cactus V2 vs a Plus II.

If this is not about OCF, then likely somene should perhaps change this thread to OT, since it is not at the time that I am posting this comment. :-)
Alfredk PRO 5 years ago
If your opinion pleases you, so be it!
thorsten198 Posted 5 years ago. Edited by thorsten198 (member) 5 years ago
@ Alfred

It should read - jump to the left - I was in a rush - sorry

I think I made clear, that I only described my personal impression, based on the camera histogram, I never claimed that my personal observation would have scientific relevance, in fact I’m not able to provide a solid, measured test in this respect – I count on rangfindergeneral’s announced test.

@ Jerry P.H.

Call it off topic if you want, but I'm firstly interested in the flash performance, I doubt that using triggers is helpful in this case.
The main question for me is, are Nikon flashes able to serve full power at xsync or not, and at the moment it looks rather like a - not!

regards
Thorsten
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
The plot thickens :))
I could not confirm a drop of power yesterday with off camera flash and my studio strobes.
Now, I just run a test like did and I can confirm his findings using a SB700 on camera in manual mode at full power.
My test was by no means scientific and I claim no repeatability!

There was no difference between the exposures up to 1/160, at 1/200 I had a visible difference in exposure and another jump when going from 1/200 to 1/250.

Unfortunately I will not have enough time to dig deeper in to it at this time, I am rather busy with Christmas.
Jerry P. H.:
One would see greater differences with say that Cactus V2 vs a Plus II.


So now a Plus II and a Cactus V2 comparison.. What is the latency of the Plus II's that put them amongst the fastest in that case..? 1/2000s as quoted on page 4 of their Manual? That doesn't put them amongst the fastest.

...despite that this thread is about flash duration and On Camera Flash and radios take no part other than add delay and cause a hinderance.
Commercial Photographer Posted 5 years ago. Edited by Commercial Photographer (member) 5 years ago
Alfredk:
Now, I just run a test like did and I can confirm his findings using a SB700 on camera in manual mode at full power..

..at 1/200 I had a visible difference in exposure and another jump when going from 1/200 to 1/250.


THATS a lot different to:

Alfredk:
I see no visual or measurable difference between 1/60 to 1/200 with my Einstein, SB700 and SB800 on full power.
NADA


So, SB900 does it, SB910 does it, SB800 does it, SB700 does it... all without REALLY trying.

What might that suggest?

.
Jerry P. H. Posted 5 years ago. Edited by Jerry P. H. (member) 5 years ago
thorsten198 said:
The main question for me is, are Nikon flashes able to serve full power at xsync or not, and at the moment it looks rather like a - not!
==========================

I hate to tell you, but I think that you will find that not just Nikon flashes, but any and all flashes do pretty much the very same thing.

Just tested the older Metz... identical results with the Sekonic meter.
Add the Vivitar 285HV to the list now too.

It is all very subtle and all pretty much in and around a third of a stop in my tests so far. One needs to get the meter really close to the flash to be able to see the results, though... like to get that 1/3rd of a stop to show on the meter between 1/60th and 1/250th, it is 3 feet or closer. At 5 feet or more, it just doesn't register anymore.
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
I think you should read my post again before opening your blow hole!

The first test yesterday was done off camera and there was no difference in exposures!

Your selective reading is getting on my nerves, just saying....OK!
Alfredk:
Your selective reading is getting on my nerves, just saying....OK!


I read everything. That includes the OP I am responding to including its context and application. Its a great pity you don't pay as much attention you wouldn't then make such ridiculous and inappropriate statements. If you want to go off at a tangent - like many here - and confuse things more by your ill founded contradictory findings - make sure that they're applicable.

Heres your complete post from yesterday:
Alfredk:
I am coming up with the exact same results, so I am not sure where this comes from?

I see no visual or measurable difference between 1/60 to 1/200 with my Einstein, SB700 and SB800 on full power.
NADA

I do know from earlier tests that I start losing exposure when using my AB B1600 on low power and a speed of 1/180 and above.

I did all tests with my D7000


The "Off camera" reference was where exactly?

Just saying :-)
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
Sorry you can't concentrate when reading, my first line in a previous comment started with this....

The plot thickens :))
I could not confirm a drop of power yesterday with off camera flash and my studio strobes.


Plus if you employ your brain it will be clear to you that I could not have used a Einstein on camera!

I think I had enough of your stupid arguments, you are starting to remind me of someone who used to hang around here, if you have a personal issue with me, send me a FM!

Go take some pictures!
MOD
Nionyn_ PRO 5 years ago
Alfredk PRO 5 years ago
Thanks!
MOD
Nionyn_ PRO 5 years ago
;-)
mjkzz (a.k.a zwdeal) Posted 5 years ago. Edited by mjkzz (a.k.a zwdeal) (member) 5 years ago
Commercial Photographer:

Come to think of it, if a meter is used, ie, you shoot at meter and get readings for different speeds, the trigger can be used to trigger the flash without instroducing any problem -- meter (in un-corded mode) starts sampling when it sees flash and end sampling when it reaches preset time interval, so no sync, no delay issue here.

Of course this opens up another can of worm, maybe another topic, that, when flash duration is close to shutter openning interval (hence speed), the meter might give a reading that is not what exposure camera will record due to trigger latency.

For example, if flash duration is 4ms, you set shutter speed at 1/250 (exposure time of 4ms) , a delay of 1ms by a radio trigger will cause actually exposure time to be 3ms, less than a meter sees, thus some might blame the meter not givng accurate reading.

This is where a FAST strobe like E640 is beneficial because its duration is so short that this is not a problem anymore.

just 2 c
rudy__ 5 years ago
I'm so glad that Canon flashes do not have this problem. ;)

I wish my photography was good enough that 1/3 of a stop variation when shooting at full power at max sync speed would make a difference. I'm just thrilled if I get the shot framed right, the subject in focus, and capture the right moment.

And I did test it with a 580EX and I see the same minor issue. Of course it isn't as bad as with Nikon. ;)
Murray McMaster 5 years ago
Regarding metering, (Alfred, I'm not doing this to piss you off) use the white towel method outlined here forums.dpreview.com/forums/post/20168236
Get the spike close to the right at 1/60 and it should move to the left as shutter speed duration is shortened when using a single flash at full power according to the results I've read here today.
Anyone want to try 2 of the same model flash at half power vs 1 at full? I am interested to see the results of that. I'm not at home with my sb600s for a day or 2 to try myself.
Alfredk PRO 5 years ago
hehehe why would I get pissed off at you, you make a perfectly valid point!

I rarely ever get pissed off, I get annoyed at people who ..............:))
Rangefindergeneral 5 years ago
Déjà vu....



I'll try and test this, but time might be against me..
I must say I've noticed a drop off in power before now but I found it happening at 1/800sec with the D70 but these weren't proper tests..
Just got to remember to take the D70....
Mr. Speedlight Posted 5 years ago. Edited by Mr. Speedlight (member) 5 years ago
I've lost my faith inlighting.

My life has been a totallie.

How did I ever manage.


Don, if this is in response to my call for a moment of silence for Canon photographers please know that I was talking trash and knew it.

Dave

---

For those shooting white brick wall do place you exposure in the middle tones as placing the exposure in the highlights will hide the exposure loss due to compression.

For those who don't see a difference try jumping from 1/250 to 1/60 and place 1/250 left of center and let 1/60 fall just right of center. If you don't see it that way then it's the work of the devil.

Dave
John Adkins II PRO Posted 5 years ago. Edited by John Adkins II (member) 5 years ago
Ok, here's my round 2 testing, with the flash on camera, and at full power. If you click through to the photo you can see all of the camera settings, but the first four are:

f/18 @ 1/60 sec ISO 200
f/18 @ 1/125 sec ISO 200
f/18 @ 1/250 sec ISO 200
f/18 @ 1/500 sec ISO 200

Light Loss Test Round 2

So I noticed that there was no light loss going from 1/60 to 1/125 and a slight light loss going from 1/125 to 1/250, but long story short... its not significant enough to ruffle my tail feathers, barely even noticeable in my opinion.

The light loss from 1/250 sec to 1/500 sec is clearly noticeable because I went over the camera's flash sync speed, but what I find even more interesting are all of the shots after that. I'm not posting all of the settings here (you can click on the photo to see them) but all I can say is that last one is f/4.5 @ 1/8000 sec ISO 1600... which I find incredibly cool.

These are all straight out of camera, except for a slight crop and text added.

Summary, on the Nikon SB-800, the light loss is not noticeable enough to worry about when not exceeding the camera's flash sync speed ...imho.

Oh, and if anyone would like to see an original photo with EXIF in tact, please say so, I have all of them
John Adkins II PRO 5 years ago
Oh, and I added this to the group pool, even though I know its most definitely not off camera flash, so feel free to nuke it from the pool, I just thought it might be helpful.
Jerry P. H. 5 years ago
Hey John, there is no need for me to post my results, they would be pretty much near identical to yours, except that when using the SB-900, the loss seems to be ever so slightly less than what your SB-800 series shows, and this also coincides with what my Sekonic meter read.

Just for S&G, I did tests on and off camera using triggers and saw no changes in the histograms, they are impossible to tell apart without EXIF, so that kinda covers both sides nicely too.

Again, good job. :-)
John Adkins II PRO 5 years ago
Jerry P. H.:

Thanks, and also good to know on the triggers, I didn't try this using remotes!
MOD
Nionyn_ PRO Posted 5 years ago. Edited by Nionyn_ (moderator) 5 years ago
John Adkins II: I won't be kicking that from the pool (though I suppose it's possible another mod might do so). It's a very useful and practical demonstration that is totally relevant to the group even if you didn't use off-camera flash for it.
Thanks for going to the time and trouble of doing the legwork! It's very much appreciated! :-)

EDIT: Beats me how you managed to strike the same pose and expression for every individual pic!
John Adkins II PRO 5 years ago
Nionyn_:

Lol, that's my 'how I look all the time' look ;)
Alfredk PRO Posted 5 years ago. Edited by Alfredk (member) 5 years ago
Cool test John!
Thanks!

I agree,, it's not worth fuzzing about the small change in exposure.
Alfredk PRO 5 years ago
Jerry, am I reading this correct? I understand that you had no variation in exposure when using the flash off camera?
If this is the case, that would confirm my crude tests from yesterday.
thorsten198 5 years ago
John Adkins II

Many thanks John - very appreciated
Mr. Speedlight Posted 5 years ago. Edited by Mr. Speedlight (member) 5 years ago
I used a D300s with a Nikon SB-700 on full power. The SB-700 was in the hot shoe. The lens was a 35-70/2.8D AF Nikkor set at 50mm, f10. The shutter speeds run from 1/320th down to 1/60 as can be read in these quick LCD photos with a hand held P&S. If you don't see a change in exposure particularly from a 1/320 to a 1/250 and again from a 1/250 to a 1/200 I can't help you. There isn't much to see from a 1/200 down to a 1/160. I don't think this is worth loosing sleep over.



The SB-800 shows the same loss at the higher sync speeds but it's a trifle more pronounced. I would expect about the same results from the SB-900/910.

This would be more important if one were shooting transparency film. The F5 offered an optional 1/300 sync speed and noted a limitation in GN or something.

That's all folks!

Dave Hartman
mjkzz (a.k.a zwdeal) Posted 5 years ago. Edited by mjkzz (a.k.a zwdeal) (member) 5 years ago
Mr. Speedlight:

Two thumbs up! Clear and concise evidence. Kowtows. Learnt a new trick, too. Thanks.
thorsten198 5 years ago
Wow Dave!

The screenshots are made with a P&S?- incredible!
Commercial Photographer Posted 5 years ago. Edited by Commercial Photographer (member) 5 years ago
Alfredk:
Sorry you can't concentrate when reading, my first line in a previous comment started with this....

The plot thickens :))
I could not confirm a drop of power yesterday with off camera flash and my studio strobes.


That only goes to prove it takes you two days to understand exactly what the OP was all about. If you go in enough circles you'll eventually arrive at an appropriate answer AFTER dissing everybody else's input brought about by your own confusion.

Reading and understanding the OP would help you immensely.

You have heard of 'Sync Cables' haven't you?
Sync Cables They're extremely useful if you don't want to use any radio or optical link - especially with an Einstein and the problem with fitting it to your hotshoe. BTW: An Einstein ISN'T a Speedlight - Thats a TANGENT.

Oh, just read this. Might as well point it out now:
Alfredk:
Jerry, am I reading this correct? I understand that you had no variation in exposure when using the flash off camera?
If this is the case, that would confirm my crude tests from yesterday.


Now you're confirming your "at 1/200 I had a visible difference in exposure and another jump when going from 1/200 to 1/250." AND "I see no visual or measurable difference between 1/60 to 1/200" ...with a practically HEARSAY test saying BOTH were the same and contradicting what you even tried and found yourself !??

FFS - Incredulous, fantastic and simply unbelievable !!
John Adkins II:
Ok, here's my round 2 testing, with the flash on camera, and at full power. If you click through to the photo you can see all of the camera settings, but the first four are:

f/18 @ 1/60 sec ISO 200
f/18 @ 1/125 sec ISO 200
f/18 @ 1/250 sec ISO 200
f/18 @ 1/500 sec ISO 200
Light Loss Test Round 2 by John Adkins II

Oh, and if anyone would like to see an original photo with EXIF in tact, please say so, I have all of them


Incredibly your variation is 5%. Exactly the same as your CLS version. Do you have some Active D-light setting set or some other highlight retention tool operating in camera or are you using rear curtain sync?

I'd like to see the EXIF (general) as your results are unreal and I think you have something set wrong which is skewing your result.
Commercial Photographer Posted 5 years ago. Edited by Commercial Photographer (member) 5 years ago
Mr. Speedlight:
If you don't see a change in exposure particularly from a 1/320 to a 1/250 and again from a 1/250 to a 1/200 I can't help you. There isn't much to see from a 1/200 down to a 1/160. I don't think this is worth loosing sleep over.

The SB-800 shows the same loss at the higher sync speeds but it's a trifle more pronounced. I would expect about the same results from the SB-900/910.


The interesting comparison is the 1/60 to 1/250 which shows a clear loss IMO:


BUT, of equal interest is the failure at 1/320s to retain the same level of flash benefit that was achieved at 1/250s. (Again, the shutter duration causing havoc with the flash duration). Does the D300s have the same 1/320FP setting as the D7000?

If the D300s has the 1/320 FP setting.. Is that loss equal to the 1/3 stop increase in shutter speed or is it greater? It looks considerably more than the 1/60-1/250s loss and you might find using this flash and camera combination more beneficial at just the 1/250s than the 1/320s when you need the maximum flash benefit.
zwdeal:
Come to think of it, if a meter is used, ie, you shoot at meter and get readings for different speeds, the trigger can be used to trigger the flash without instroducing any problem -- meter (in un-corded mode) starts sampling when it sees flash and end sampling when it reaches preset time interval, so no sync, no delay issue here.


A useful point as usual and one worth remembering when metering.

Using the appropriate setting would be the most reliable.. Corded, Non-Cord or Radio (via cord?) to suit the application.

On the other hand, a flashmeter with a built-in radio link would be adversely affected if trying to check output results by readings in this, the OP's scenario, unless they're used non-corded.
mjkzz (a.k.a zwdeal) 5 years ago
Commercial Photographer:

I am not sure how corded mode works, never used it but I do know the uncorded mode only starts sampling when it sees flash light.

On the other hand, even if it is corded, I am not sure how it operates -- does it start sampling when it sends out fire signal to flash(es) or does it wait for light from flash even though it has commanded flash(es) to fire.

In the first case, the delay in radio signal could cause mis-reading (or right reading because the camera will behave the same way) .

In the later case, the delay by radio does not matter because it is the flash light that starts the sampling.

To further learn this, I am going to construct a circuit to inert a small delay, like 10ms, to the corded mode and see how the meter reacts.

Of course, another thing is, even if I find out how my Sekonic L-308S works, it does not mean all other meters work that way.

Just 2 c
zwdeal:
I am not sure how corded mode works, never used it but I do know the uncorded mode only starts sampling when it sees flash light.


Corded mode is simply plugging a sync cord into the flash meter and triggering via a button on the flash meter. You could plug a radio here and introduce an additional delay if you wanted.

Exactly what the delay is would probably be different for every meter available.

It would be interesting to know how your L-308S works.
1
(1 to 100 of 157 replies)
Groups Beta